Info from John Carmack (4)

Bernd Kreimeier (Bernd.Kreimeier@NeRo.Uni-Bonn.DE)
Mon, 3 Jun 1996 19:24:50 +0200 (MET DST)

Date: Mon, 3 Jun 1996 19:24:50 +0200 (MET DST)
From: Bernd Kreimeier <Bernd.Kreimeier@NeRo.Uni-Bonn.DE>
Message-Id: <>
Subject: Info from John Carmack (4)

----- Begin Included Message -----

>From Sat May 11 05:46 MET 1996
Mime-Version: 1.0 (NeXT Mail 3.3risc v118.3)
Content-Transfer-Encoding: quoted-printable
X-Nextstep-Mailer: Mail 3.3 (Enhance 1.0)
From: John Carmack <>
Date: Fri, 10 May 96 22:46:23 -0500
To: Bernd Kreimeier <Bernd.Kreimeier@NeRo.Uni-Bonn.DE>
Subject: Re: q&a (2)
Content-Type: text/plain; charset="us-ascii"
Content-Length: 2727

You wrote:
> 1. The Quake registered (==shareware?) will still be released as
> DJGPP-linkeable modules?

The initial 0.9 will not, but some later version will.

I'm not sure if the initial CD pressing will, because there won't
be any external drivers yet.

The will be net released at some point for sure.

> Followup: Iff, will the linux version be on the same CDROM? As
> modules (gcc-linkeable) as well? I would hate to see DOS get the
> better of this. Or looseDOWS.

I have intentions to do so, but since Dave Talor left id, there is
nobody here actively running linux, so it can get pushed out of the
way by other issues pretty easily. Dave is still going to support
the unix ports for us, but it is less of a priority.

Yes, unix versions are cool, but they don't rank that high on the
priority list.

> 2. The COLORMAP remapping of the WAD2 palette for lighting (32
> steps again?) is part of the EXE now?

Its an external data lump now, so the palette and colormapping
(including fullbrighting) can be completely changed.

> Followup: like the normal vectors used with the MDL, wouldn't it be
> a better idea to keep any lookups outside the EXE, to prevent EXE
> hacking, and ease configuration?


> 3. Is the design of QuakeC vs. physics of the game flexible enough
> to separate "blast" (momentum transfer) from "damage", i.e. pushing
> an armored target without hurting it (area effects, not direct
> hits)?

Absolutely. Romero has asked me to do this a number of times, but
I haven't gotten around to it. A while ago, we had mosnter cannons
that chucked bad guys out of a piston, but it went away with some

> Followup: will involuntary movements (like being pushed by an
> explosion) change the view_up (including roll)? Whatever
> restrictions on GUI/net packet size, preserving the view_up during
> involuntary movements (falling) seems to make things "stiff" by
> design.

I let the client short curcuit the view angle in most cases, so the
server doesn't usually override it. It's possible, though.

> 4. Just a confirmation: the z-fill is done as a separate pass, not
> within the texture mapping loop, because of (lack of) registers on
> i586?


> Better performance on other CPU's? Alpha?

Unlikely, because we have a fair amount of hand coded intel ASM.
A 21164 certainly won't be a slug, though. I still need to fix a
little 64 bitism in my code before it is going to work on alphas,
though. I will get that done for sure.

John Carmack

----- End Included Message -----